MR DWI does not substitute for stroke severity scores in predicting stroke outcome

Abstract

The accurate prediction of outcome following stroke is of great importance. In the clinical arena, accurate outcome prediction can enable physicians to be involved in informed discussions with the patient and caregivers, to accurately define treatment goals, to objectively balance the risks and benefits of treatment options, and to make early decisions about later discharge. There also has been great interest in applying predictive models to the design and analysis of acute stroke treatment trials. Predictors measured at baseline, such as the initial NIH Stroke Scale (NIHSS), can identify patients with good, intermediate, or poor prognosis if untreated. Trials will then enroll only those with intermediate prognosis under the assumption that these are most likely to show benefit.1–3 Although logical and attractive, the validity of this approach remains to be proven.2 Furthermore, a priori restriction of eligibility by baseline predictors of outcome could result in the failure to detect a benefit that is present in all patients. The efficacy of IV tissue plasminogen activator (tPA) is not affected by baseline NIHSS.4 More recently, the baseline NIHSS has been used to define the criteria for a beneficial clinical outcome.3 Instead of selecting a single criterion for all patients, such as a modified Rankin Scale score (mRS) of 0 to 1, the criterion for those with a higher baseline NIHSS and a worse prognosis is liberalized to include mRS 0 to 2. Whether this approach …